Extração e Análise de Dados de Rádio FM: Utilizando SDR com Inteligência Artificial

• Thay Guimarães Bellona IFAC
• Davi Brito Do Nascimento IFAC
• Flávio Miranda de Farias IFAC
• Debora C. Muchaluat-Saade UFF

DOI: https://doi.org/10.5753/webmedia_estendido.2026.34493

Resumo

Radio frequency monitoring via Software Defined Radio (SDR) constitutes a relevant source of information in Open Source Intelligence (OSINT) and public auditing scenarios. However, the continuous analysis of analog audio generates significant operational and computational bottlenecks. This work presents the development of an open-source software tool based on the edge computing paradigm. The application integrates low-cost RTL-SDR receivers to a native Python graphical interface (PyQt6), local Artificial Intelligence models for Automatic Speech Recognition (Whisper), and Natural Language Processing (NLP) modules. The solution automates the acquisition, signal demodulation, continuous transcription, and quantitative extraction of discursive patterns through N-grams and ontological mapping. Preliminary operational tests conducted over an 8.5-hour continuous window indicate the feasibility of converting electromagnetic signals into structured databases, suggesting the potential for adopting radio in the ’audio as data’ analytical paradigm.

Palavras-chave: Software Defined Radio, Artificial Intelligence, Automatic Speech Recognition, Text Mining, Open Source Intelligence

Referências

Ricardo Baeza-Yates and Berthier Ribeiro-Neto. 2013. Recuperação de informação: conceitos e tecnologia nas máquinas de busca (2nd ed.). Bookman Editora, Porto Alegre, Brasil.

Grady Booch. 2005. The unified modeling language user guide (2nd ed.). Pearson Education India, Upper Saddle River, NJ, USA.

Ronaldo Goldschmidt, Emmanuel Passos, and Eduardo Bezerra. 2015. Data Mining: conceitos, técnicas, algoritmos, orientações e aplicações (2nd ed.). Elsevier Brasil, Rio de Janeiro, Brasil.

Daniel Jurafsky and James H. Martin. 2024. Speech and Language Processing (3rd ed.). Prentice Hall, Upper Saddle River, NJ, USA.

Kantar IBOPE Media. 2023. Inside Áudio: O Consumo de Rádio no Brasil. Relatório de audiência e alcance de rádio-frequência. Acessado em: 25 maio 2026.

Anass Khayati, Chaïmaâ Kissi, and Adnane Addaim. 2025. Design and Implementation of a Software Defined Radio Reception Chain for Automatic Identification System (AIS). In IEEE International Conference on Advanced Communication Technologies. IEEE, Piscataway, NJ, USA, 1–6.

Dean Knox and Christopher Lucas. 2021. Potential and Pitfalls of Audio as Data for Political Research. Political Analysis 29, 4 (2021), 381–404.

Marc Lichtman. 2020. PySDR: A Guide to SDR and DSP using Python. [link] Acessado em: Abril de 2026.

Christopher D. Manning and Hinrich Schütze. 1999. Foundations of statistical natural language processing. MIT Press, Cambridge, MA, USA.

Rafael Mestre and Matt Ryan. 2026. Potential and Pitfalls of Audio as Data for Political Research: Alignment, Features, and Classification Models. Political Analysis (2026).

Alan V. Oppenheim, Ronald W. Schafer, and John R. Buck. 1999. Discrete-Time Signal Processing (2nd ed.). Prentice Hall, Upper Saddle River, NJ, USA.

Roger S. Pressman and Bruce R. Maxim. 2021. Engenharia de software - 9ª Edição. McGraw Hill Brasil, Porto Alegre, Brasil.

W. Promteerawong and M. Ketcham. 2020. The System for Storing and Analyzing FM Broadcasting Content. International Journal of Networked and Distributed Computing 8, 2 (2020), 85–90.

Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2022. Robust Speech Recognition via Large-Scale Weak Supervision. arXiv preprint arXiv:2212.04356 (2022).

Ian Sommervile. 2019. Engenharia de Software (10th ed.). Pearson, São Paulo, Brasil.

Tribunal Superior Eleitoral. 2019. Resolução nº 23.610, de 18 de dezembro de 2019. [link] Dispõe sobre propaganda eleitoral, utilização e geração do horário gratuito e condutas ilícitas em campanha eleitoral. Acessado em: 02 ago. 2026.

Raul Sidnei Wazlawick. 2014. Metodologia de Pesquisa para Ciência da Computação (2nd ed.). Elsevier Brasil, Rio de Janeiro, Brasil.

Jia Zhang, Yinian Zhou, Rui Xi, Shuai Li, Junchen Guo, and Yuan He. 2022. AmbiEar: mmWave Based Voice Recognition in NLoS Scenarios. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 6, 2 (2022), 1–25.

Tan Zhang, A. Vance, Y. Cui, and S. Banerjee. 2014. Towards Commoditized Real-time Spectrum Monitoring. In Proceedings of the 13th ACM Workshop on Hot Topics in Networks (HotNets). ACM, New York, NY, USA, 1–7.

Azeddine Zidouni, Sophie Rosset, and P. Zweigenbaum. 2010. Efficient combined approach for named entity recognition in spoken language. In ISCA Archive. 633–643.